Journal of Scientific & Industrial Research Vol. 61. February 2002. pp 128- 138

Effect of Experimental Variables on Phenol Adsorption on Activated Carbon Prepared from Coconut Husk by Single-step Steam Pyrolysis: Mass Transfer

Process and Equilibrium Studies

v P Vinod and T S Anirudhan*

Department of Chemistry. University of Kerala. Kari avattom. Trivandrum 695 58 1. India

Received: 14 September 200 I: accepted:07 November 200 I

Industrial solid wastc can be convcrted to activated carbon and is used as low cost adsorbent for wastewater treatment. Coconut husk, a waste product in coil' industry was used for the preparation of activated carbon by single-step steam pyrolysis. The product exhibits a very hi gh adsorption potential for phenol and under optimum condi ti on more than 99 per cent removal has been achieved . The adsorption of phenol onto activated carbon has been sllldi ed. using batch tech niqu e under kinetic and equ ilibrium conditions. The effect of several va ri ab les such as initi al concentration. temperature. agi tation speed. particle size of the adsorbent and pH on phenol remova l has been studied. Sorption of phenol depends on the pH of the solution with max imum uptake of 99 per cent occurring at 5.5 and an initi al conc. of 25.0 mg/L. Both. ex ternal and il1!raparticular. mass transfer diffu sion models have been tested to describe the kinetic data . The sorpti on mechani sm is al so disc ussed. The equilibrium data have been evaluated for compli ance with Langmuir and Freundl ich isotherm models. Temperature sllIdy reveals that the adsorption of phenol is exothermic in nature. Thermodynamic parameters have also been presented to predict the nature of adsorption. Isosteric heat of adsorption is varied with surface coverage. The adsorpt ion data show th at spent carbon can be regcnerated for further use by hot water treatment.

Introduction

Di scharge of phenoli c wastes imparts a carbolic odour to water body and has detrimental effects on human be in gs and fish. The source of phenolic pollution is from the discharge of various industri es such as petroleum-refinery, pharmaceuticals , paints , steel, petrochemical, and plywood I. Environmental Protecti on Agency (EPA ) has set a limit of 0.1 mg/L or phenol in wastewater. According to the Indi an Standard Instituti on2

, the permissibl e limit for phenol for the discharge of effluents into inl and surface water is 1.0 mg/L. In the pas t two decades, several . tudics have been carried out on adsorption, as an efficient and economica ll y feasible process, for removing phenols from aqueous industrial effluents. Many researchers have shown that activated carbon s prepared from industrial and agr icu ltural by-products or was te materials is an effective adsorbent for the removal of phenol from wastewaters. Cost­effec tiveness , availability, and adsorption properties are the main criteria for choos ing an adsorbent to

* Aut hor for corresponde nce IE-mai l: tsani @redilTmail.coml

remove pollutants. Taking these criteri a into cons ideration , earli er workers used the activated carbon prepared from jute stick'" sawdus t'), Sarkanda grass5

, palm tree cobs6, rice husk?, tamarind nut8

,

Moringa oleifera husk9, Coconut coir lO

, and Salvinia molesta Mitchell ".

A two-stage carbonization-act ivati on process, in volving heating the precursor to a particular temperature under nitrogen, fo llowed by ex posure to a steam/nitrogen mixture at higher temperature is usually used for the production of activa ted carbon rrom vanous materials. This process requires nitrogen , which is not easily available in many parts of the developing countri es like Indi a, and two heating stages , involving hi gh-energy in put. Earli er workers 12, demon strated tha t single-step steam pyrol ys is activation produced hi gh quali ty carbon for wastewater treatment. In the present study we use single-step steam pyrolysis for producing act ivated carbon because of its advantages such as:( i) It req utres no chemical input. (ii ) It requires less energy than traditional processes wh ich involving two heati ng steps, frequentl y a hi gh temperature, and (ii i) Local availabil ity of low-cost carbon by thi s

VINOD & AN IRUDHA el al.: PH ENOL ADSORPTION ON ACTIVATED CARBON-A STUDY 129

technique may encourage polluting industries to install treatment pl ant. Coconut husk from coir industry is considered as a waste material , since it does not find any use as such and causes di sposal problem. Manju e l al. 13 have studi ed the ability of coconut hu sk-based carbon to immobili ze arsenic from wastewaters. Kin eti c and equilibrium studi es in adsorpti on give information about the mechani sm of the process and adsorption capacity of the adsorbent and also essenti al to design an industrial adsorber. The present work aims to study the influence of ex perimental vari abl es in phenol adsorption under kinetic and equilibrium conditions on activated carbon prepared from coconut husk by single-step steam pyrolys is.

Experimental Procedure

The activated carbon used in thi s work was prepared from coconut hu sk (CH). CH was obtained from a loca l coil' industry. It was grou nd , washed with distilled water to remove dirt and fine particles « 0.096 mm), and dried at SOoC for 4 h. About 50 g of CH was placed in a purpose-made graphite tube and placed at the center of the Matri (Indi a) made furn ace. Steam, produced by a steam generator, entered the graphite tube at a rate of approx. 3 mLimin (liquid water). The sample was then heated at 10°C/m in to 600"C and hel d at that temperature for I h. After allowing the furnace to cool to room temperature the carbonized materi al was washed with 0.1 M HCI and then with di stilled water. The washed product was dried in an oven at 100°C until constant weight and cooled. The yield of the acti vated carbon was fou nd to be 42 per cent of the ori gin al materia l. The adsorbent was ground and the partic les having the average diam 0.096 mm (-SO+230mesh) was collected and kept in desiccator for subsequent use.

The stand ard methods " were used to determ ine the surface and physical properties of the acti vated carbon. The surface area of the ca rbon was determ ined by the N2 BET method, us ing a Quantasorb surface analyzer (model-0517) and porosity by mercury porosimeter. The zero point charge, pH ,pc is defined as the pH of the suspension at which the surface charge density Go = O. A potentiometric titrat ion method was used to determine Go (C/c m2

) as the funct ion of pH and ionic strength . The Go was ca lculated from the titre value using Eq . ( 1)14 .

Go F (CA -CB +[OH- ] - [H +])

A .. . ( I )

where F is the Faraday constant, A is the surface area of the suspension (c m2/L ), C" and Co are the concentration of ac id and base (eq/L) after each addition during the titrati on and [OH-] and [H +J are equivalents of the OH- ~lI1d H+ ions bound to the suspension surface (eq/c m\ The point of intersection of the Go vs pH curve gives pHz!'r' The carbon was ignited in a muffle furn ace at 100°C for about 3 h to determine the moi sture content. The characteristics of the activated carbon are: surface area, 421 m2/g; porosity, 0.51 mLlg; pHzp", 3.45 ; apparent density. 1.2 g/mL; ani on exchange capacity 0.S3 meq/g; moisture content, 6 per cent ; ash content , 2.3 per cent and iodine number 579. FTTR spect ra of activated carbon were recorded on a Perkin Elmer FTfR spectrometer (model-I SO) between 400-4000 cm-I. TR spectra of carbon showed the presence of conjugated hydrogen bonded carbonyl group (band at 16 13 cn,- I). as sugges ted by Hallum and Drushel l

). The additional peaks at 1750 cm-I (ur=,,) and 1450 cm-I (uc-o) indicate the presence of carboxyl ic (-COOH) groups on the carbon su rface 16.

The equilibrium isotherms were determined by equilibrat ing 2.0 giL 0" ac tivated carbon with 50 mL of phenol solu tion in a glass stoppered bottl e of 100 mL capac ity. These were shaken in a temperature con trol led water bath shaker (Remi model G-1 6) at 200 rpm for 4 h. Aliquots of phenol so lu ti ons were withdrawn at different preselected reaction time and filtered. The filtrate was analysed for phenol content , using uv-visible pectrophotometer at 270 nm . Once the residual concentration (C, mg/Ll is known. the phenol concentration in the carbon (qt , mg/g ) was determined using the mass ba lance formul a qt = (Co-Ce) VIlli. where Co is the initial concent ration (mg/L), V is the vol ume of the solu tion (mL) and 111 is the mass of the adsorbent (g) .

Several contact time experiments were carried out at constant pH to study the effect of the num ber of' va ri ables such as the effect of initi a l concent rati on of' phenol , temperature, ag itat ion speed and particle size of the adsorbent on ad sorption . Effec t of pH was studied by adjust in g the initial pH of phenol soiution to different va lues between 2 and 10 with di!. aOH and HC!. The adsorpti on isotherms at different 10-40°C were obtai ned at pH 5.5 and fixed adsorbent cone. of 2 giL for va ryi ng concentrati on of phenol.

i30 J SCI IND RES VOL 61 FEBRUARY 2002

ranging fro m 50 to 500 mg/L. In these studies the test so lut ions were ag itated until equilibrium had been at tained. After atta inment of equ ilibrium, the supernatant liquid was carefull y filtered and desorp tion experi ments were carried out uSll1 g distilled water and hot water. To regenerate the adsorbent the adsorption and desorption process were foliowed for three cycles. After each desorption C" cles the spent adsorbent was washed with disti li ed water to remove unadsorbed phenol from carbon surfa e and dried at room temperature. Experi ments were carried out in duplicate and the per centage error

was < 5.0.

Results and Discussion

Figur 1 (a) presents the effect of agitation time on the adsorption of phenol for different initial concentrations . The curves shows strong adsorption during 30 min and equil ibrium is establi shed in appro x. 4 h. The saturation peri od of the adsorpti on is entirely independent of the in itia l concentration . With an increase in initial concentration of phenol from 50 to 125 mg/L. the uptake of phenol decreases fro m 94.80 per cent (23.7 mg/g) to 72 . J 6 per cent (45. 1 mg/g). The resu lts show that the percentage removaJ I:" Influenced by the initi al concentration. Th is is because at high initia l concentration the ratio of the II1l tial number of moles of phenol to the availab le urface area is h igh. hence the fractional adsorpt ion

becomes dependent on initia l concentrati on.

Figure 1 (b) shows a seri es of agitati on time curves a t 10, 20, 30 and 40 °C respectively. The percentage adsorpti on of phenol decreased from 98.0 per cent (24.5 mg/g) to 86.6 per cent (2 1.65 mg/g) with increase of temperature from 10 to 40°C. This ind icates that the lower the temperature the greater will be the adsorpt ion. Figure 1 (c) shows that the amount of phenol sorbed from solu t ion increases wi th the lapse of time and reaches the saturation time in 4 h. The saturation of adsorption is enti re ly independent of agitation rate as confirmed by the results presented in Figure 1 (c) . The max imum adsorpt ion of 96.0 pe r cent (23 .99 mg/g) is observed at an ag itati on rate of 400 rpm and it decreases to 80.0 per cent (20.0 mg/g) when the agitat ion rate decreases to 100 rpm . The results indi cate that ex terna l adsorpt ion of phenol onto carbon is controlled by the degree of agitati on. As can be seen from Figure 1 (d) the rate of adsorpti on is in f1u e nced by the partic le s ize of ca rbon . The removal of phenol

Time (min)

rigure I-The e ffeet 0(" Ca) initi al coneentration. (b) temperature. (e) agi tat ion rate and (d ) panicle size 01 the adsorhent on the kinet ics of phenol adsorp ti on by ac ti vated carbon

VINOD & AN IRUDHAN el al. : PHENOL ADSORPTION ON ACTIVATED CARBON-A STUDY 131

by carbon decreases from 94.4 per cent (23.6 mg/g) to 77.96 per cent ( 19.49 mg/g) by increas ing the particle size fro m 7.4 x 10-3 to 50.8 X 10-3 cm. The higher removai with the sma lle r particle has been attri buted to greater access to the internal pores. i.e., shorter path lengths and to the large surface area per unit weight of ca rbon.

The two important aspects for parameter eval uation of the adsorption study are the kineti c and the equilibria of adsorption . The kinetic constants of phenol adsorption, which would be used to optimize the res idua l time of industri al wastewaters in acti vated carbon column , are measured at different experimenta l variables. For the porous adsorbent stud ied, sorption ki netics are mai nl y control led by three stages including diffusion process i 7

. The first tage involves the transport of the adsorbate to the

external surface of the adsorbent. The econd stage is the d iffusion of the sorbate in to the pores of the adsorbent. then proceeds and it is finally adsorbed into the intern al surface of the adsorbent. The last stage is re la tive ly rap id and non- limiting phase. First stage describes extern al or film mass transfer res istance and second stage is related to the intrapart ic ul ar mass transfer diffusion model.

Due to the porous nature of the adsorbent , intraparti c le diffusion is expected to be the rate Ilmltlll g step. Hence. app lying pore di ffusion phenomena the sorption kinetics are model ised. accord ing to the fo ll ow ing intraparti cular diffusion equat ion 18 rEq. (2)] ,

where q, and qe are the adsorbate concentration of the sol id at time t and at equilibrium, respectively, d is the parti c le d iam. and D; is the diffusion coefficient In the so lid (m2/s). The dirrus ion coeffic ient of adsorption at different variab les are determined from the s lopes of the straight line plots of logrl-(q/Qc)21 vs I (Figure 2) and are g iven in Table . The va lues of D; are fo und to increase with the increasing initi a l concentration. These results are consistent with previous studies on phenol adsorption onto activated carbon I I and meta l adsorpti on onto chilosan 19 . Increasing the solute concentrati on in the so luti on seems to reduce the diffusion of so lute in the

Or----~----r-----r--____, (d)

-0.4

-0.8

-1.2

-1.6

Sorbent dose : :2 gil pH : 5.5 Temperature : 30"C Agitation speed : 200 rpm Initial conen : 50 mglL

. : 7.4 X 10.3 em

. : 10.6 X IO·Jem ~ : 14.9X lO·Jcm • : 50.8 X 10·3em

• •

Figure 2- Urano and Taehikawa plots for the adsorption of phenol on acti vated carbon at d ifferent (a) init ial conce·ntrati on. (b) temperature, (c) agitati on rate and (e1 ) parti cle size of the " ...J ........ _1...,,~.

132 J SCII NO RES VOL 6 1 FEBRUARY 2002

Table 1- Kineti c parameters for the adsorption of phenol on activated carbon.

Oi ffu sion constant Mass transfer coefficienl Vari ahles

Oi (m2/s)

Inil ial concentration(mglL)

50 2.77 x 10.1.1

75 3.03 x 10·1.1

100 3.38 x I 0.1.1

125 4.03 x I 0.1.1

Temperature eC)

10 5.29x I 0.1.1

20 5.0 Ix I0·1.1

30 2.77x I0·1.1

40 2.4 lx I0·u

Panicle size (c m)

7.4 X 10·.1 7.93 x I 0.13

10.6 X 10·.1 5.33 x 10·1.1

14.9 X 10·.1 2.47 x IO·"

50.8 X 10·.1 1.1 8 x 10.1.1

Agitat ion rate (rpm)

100 2.12x 10.1.1

200 2.77x 10.13

300 3.53x I 0.1.1

400 4.07x 10.13

boundary layer and to enhance the di ffus ion in the solid . I f intraparti cle or pore diffusion is to be the rate limit ing step the pore di ffusion coefficient should be in the range of 10. 12 to IO· I ~ m2/s20

. Here the va lues of Dj arc in t he order of 10. 13 m2/s, indicating that the rate li mi tin g step appears to be int raparti cular di ffusion for phenol-carbon system.

I t is evident from the Table I that the va lues of D, decrease w ith ri se in solution temperature. The increase in the k ineti c energy of the adsorbate species w ith the ri se of solution temperature, due to the increased relati ve escaping tendency of phenol from so lid phase to bulk phase and thus a decrease in adsorpt ion is observed. The decrease in D j with temperature indicates the exothermi c nature of the adsorpti on. The decrease in physica l force responsible for adsorpti on w ith temperature also causes the decrease in adsorpti on at higher temperatu re. The Dj

values range from 2.820 x 10.13 m2/s for an agitati on rate of 100 rpm to 4.065 x 10.13 m'!./ for 400 rpm. W ith change in particle size f rom 50.8 x 10.3 to 7.4 x 10·' cm the va lues of D j increases from 1. 184 x 10.13

to 7.932 x 10·1.1. The increase in the extend of adsorpti on by increas ing the ag itati on rate and

BL (cm/s)

0.9 159 4.40x 10.6 0.9550 0.9643 3.99x 10.6 0.9584

0.9975 2.75x I 0.6 0.9787

0.9896 2.59x 10.6 0.9945

0.999 1 5.47x 10·1> 0.9837

0.9982 4.60x 10·" 0.9688 0.9 159 4.40x 10.6 0.9550

0.9824 3.89x I0·6 0.9933

0.97 17 5.98x 10.6 0.9730

0.9642 3.7 Ix I0·6 0.9889

0.9848 2.26x I 0.6 0.9878

0.9257 1.0 1x 10.6 0.9862

0.990 1 3.66x 10.6 0.99 13

0.9 159 4.40x 10.6 0.9550

0.9587 5. 15x IO·6 0.9844

0.9369 5.96x I 0.6 0.9802

decreas ing the particle size further support the above view that intraparti cle diffusion controls the adsorpti on process.

The External mass transfer analys is of phenol during the adsorption process is studied using the mass transfer di ffus ion model developed by McKay el al. 17

•

In (~- ! J= C" 1+ lI1 kL

.. . (3)

where C, is the concentrat ion of solu te at time I, CO is the initi al concentrati on of the solute. 111 is the mass of the adsorbent per unit vo lume of pan icle-free solution of solute, /.: L. is the L angmuir constant (obtained by multi plying QO wi th b), Ss is the speci f ic surface per unit vo lume of parti cle-free slurry and 81. is the mass transfer coefficient. The va lues of 8 1. were determined from the slope and intercepts of the plots

VINOD & AN IR UD HAN et al.: PHENOL ADSORPTION ON ACTI VATED CA RBON-A STUDY 133

of In {(C/Co)-[ I/( I +mkL)]} vs f , for different experimental vari a.bles (Figure 3) using the least squares method. The values of BL are calcul ated and are presented in Table 1. From Table I, it is seen that as expected, the mass transfer coefficient suggests that the velocity of phenol transport is rapid enough to use act ivated carbon for the treatment of water and wastewater containing phenol. The BL values range from 4. 395 x 10-6 cm/s at 50 mgllL to 2 .588 x 10-6

cm/s at 125 mg/L initi al phenol concentrations . It is found that increasing the initial phenol concentrati on results in a decrease in the external mass transfer coefficient.

As the Table I shows the va lue of BL decreases from 5.467 x 10-6 at lODe to 3.888 x 10-6 cm/s at 40oe, which suggests that sorpti on is thus faster at lower temperatures. The energy of act ivati on is determined from the slope of the Arrhenius plot of In BL vs I IT (Figure not shown) and is fo und to be -7.95 kl ima I. The negati ve value of act ivati on energy sugges ts that a feasible process opt ion would comprise adsorbing the phenol onto act ivated carbon at relatively low temperature and then desorbing it at a higher temperature to regenerate the acti vated carbon. The fi ndings of the present in ves tigat ion also indicate that mass transfer of phenol is favoured at higher ag itation speed and smaller particle size of the adsorbent. The results are similar to those reported for the adsorpt ion of phenol by Salvinia molesta Mitchell-based act ivated carbonll. The effect of particle size and ag itat ion speed on mass transfer coeffic ient is poss ibl y more complex than origin all y suggested in Kolmograff' s theory due to the relati ve velocity between fluid and part icl e in the agitating system. Accord ing to the Kolmograff's theory, when the agi tat ion speed is very high the sheer force on the bound ary layer is greater which effecti ve ly red uce the thickness of the boundary film. Thus the res istance to fi lm transfer is correspondin gly reduced . The trend in BL values in Table I supports thi s concept. Smaller part icles have a larger extern al surface area availabl e fo r adsorp ti on per unit mass of adsorbent th an larger parti cles and one cou ld anti cipate hi gher BL values. The assumption that large surface area presented by smaller particles results in a lower driving force per unit area, remain very plausible l9.

Earlier studies have indicated that solution pH is an important parameter affecting the sorption of phenol on carbon surface2

1.22. In general, agricul tural or industr ial waste based activated carbons carry a

o r---~-----r----~----r---~

-I

-2

-3

-4

-5

o

-I

-2

-3

-3

(d)

Sorbent dose pH Temperature Initial eonen : 50 mgIL

(c)

Sorbent dose pH Initial conen. Temperature

(b)

:2g!L· : 5.5 : 50mgfL : 30DC

-4 Sorbent dose pH

: 2 gIL : 5.5

-5 Initial conen. : 50mgIL Agitation speed : 200 rpm

o

-I

-2

-3

-4

-5

o

(a)

Sorbent dose pH

: 2 giL : 5.5

Temperature : 30·C Agitation speed : 200 rpm

50 100

. : 7.4 X 10-1 em . : 10.6 X IO-lem ~ : J4 .9X JO-lem

150

• : 100 rpm • : 200 rpm ~ : 300 rpm • : 400 rpm

• : lOoe • : 20°C ~ : 30DC • : 400e

•

•

. : 12Smg/L ~ : JOOmg/L . : 7Smg/L . : SOmg/L

•

200 250

Time (min)

Figure 3- McKay et al. plots for the adsorpt ion of phenol on ac tivated carbon at di fferent (a) initi al concentration. (b) temperaLUre. (c)agitation rate and (d)part icle size of the ad~orbent

134 J SCII ND RES VOL 61 FEBR UA RY 2002

su rface charge, which is essentially dependent on pH of the solution. Phenol adsorpt ion by acti vated carbon as a function of solution pH is studied (Figure 4). The uptake is small with low pH range and graduall y increases up to pH 5.5 , where maximum removal of 99 and 92 per cent is observed from an ini tial concentration of 25 and 50 mg/L, respecti ve ly. At pH range greater th an 5.5 the adsorpti on capacity decreases to a minimum value at pH 10.0. Various reasons can expl ain the behaviour of the adsorbent in phenol adsorpt ion relati ve to pH. The pHzpc of the carbon is found to be 3.45. Below this pH carbon surface is pos iti ve ly charged and above the surface is negati ve. At lower pH, most of the phenol ex ists in protonated fo rm and surface of the carbon is also pos itive. Pos iti ve ly charged carbon surface repel positi ve ly charged protonated phenol and results in the reducti on of phenol adsorption. With increase in pH the molecul ar form of phenol persists in the med ium , leading to the enh ancement of phenol adsorption. It is also observed that phenol molecules are eas i Iy adsorbed compared to negat ive phenolate anions and protonated spec ies, as shown previously by others2o. Phenol adsorpti on on carbon usuall y occu rs th rough a donor-acceptor complex mechani sm involving carbonyl oxygen group on the carbon surface acting as the e lec tron donor and aromatic ring of the phenol as the acceptor. At hi gh pH values, the cont ribu ti on of OR" ions fa r exceed that of the phenolate ani on and hence occupied on the carbon surface leaving phenolate ani on unboun d. The increased solubili ty and increased hind rance to the diffusion of the pheno l ions at higher pH may also lead to a dec rease in phenol adsorption.

Adsorpti on isotherm experiments were carri ed out to eva luate the potentia l of the sorbent for commercial application. Since solution temperature has significant effect on adsorpt ion equilibrium, the phenol removal capaci ty of acti vated carbon was determined by adsorption isotherms obtained at 10, 20, 30, and 40DC, respecti vely. The shape of the isotherm gives an indicati on whether the adsorpti on is favo rable or not. The adsorpti on isotherms (qc vs Co) are regul ar, positi ve and concave to the concentrati on ax is (Figure 5) . Initi ally the adsorption is qu ite rapid , whi ch is foll owed by a slow approach to equilibrium at hi gh phenol concentration. These results indicate the ut ility of the adsorbent fo r the removal of phenol from was tewaters in a wide concentrati on range. Accordi ng to the slope of the initi al pos iti on of the

--~ • '-' c::

.S! ~ .. ~

"0 -<

100

80

60

40

20

Sorbent dose : 2 giL Temperature : 30 uC Agitation time : 4 h

Agitation speed : 200 rpm

- ; 25 mglL + : 50 mg/L

o ~--~ __ ~ __ ~ ____ ~ __ ~ __ ~ o 2 6

pH 8 10 12

Figurc 4-Thc effect of pi I on the adsorpti on or phenol onto activa tcd carbon

210

180

150

-- 120 ~ ....... ~

E 90 '-' ~

60

30

0

0 50

• : 10 °C + :20 °C ~ : 30 °C • : 40 °C

Sorbent dose : 2 giL pH : 5. 5 Agitation time : 4 h Agitation speed : 200 rpm

100 ISO 200 250 300

C. (mc/L )

Figure 5- Adsorption isotherms or phenol onto activa ted carbon

curves, isotherm may be class ified as H-type of the Gil es class ification23

. Thi s suggests that acti vated carbon has high affinity for phenol and there is no competition from the solvent fo r sorpti on sites. It is also observed from the FigureS that isotherms tend to define a plateau at higher equi librium concentrati on. Isotherms belonging to subgroupIJ of the Gi les cl ass ificati on. According to Giles et al.23

, saturati on of the surface by phenol molecules seems to be reached that is to say, a complete monolayer covering of phenol on carbon is poss ible un der used experimental condit ions.

-.~

VINOD & ANIRUDHAN el at.: PH ENOL ADSORPTION ON ACTIVATED CARBON- A STU DY 135

The isotherm data at different temperatures were processed using the Langmuir and Freundlich isotherm models. The Langmuir isotherm is based upon an assumption of monolayer adsorpti on onto a surface containing finite number of adsorption sites of uni form energies of adsorpti on with no transmigrati on of adsorbate in the plane of the surface. The linear fo rm of the Langmu ir equation is represented in Eq . (4):

... (4)

where Ce and qe are the equil ibrium adsorbate concentrat ion in liquid and solid phase, respecti vely. QO and b are Langmuir constants related to adsorption capac ity and energy of adsorption, respecti vely. The Freundli ch isotherm model assumes heterogeneous surface energies , in which energy term in the Langmuir equati on varies as a funct ion of the surface coverage due to variati on in the heat of adsorption. The linear form of the Freundlich isotherm equati on is represented by Eq . (5):

I log qe = -log Ce + 10g KF

n . .. (5)

where KF and I/n are Freundlich constants re lated to adsorpti on capac ity and intensity of adsorpti on respecti ve ly.

The plots of Celqe vs Ce gives straight lines at each temperature showing the applicability of Langmuir isotherm (Figure 6). T he Langmuir isotherm constants were computed according to the least squares fitting method, using experimental qe

and Ce values. The data are presented in Table 2. At 300 e, coconut husk based acti vated carbon adsorbed 146.16 mg/g phenol wh ich is fo und to be better adsorbent than the adsorbent such as local activated carbon (22 .0 mg/g), fi brous acti vated carbon (25 .5 mg/g), E Merck activated carbon (35 .5 mg/g) , waste ferti lizer carbon (35. 5 mg/g) , polymeric adsorbent (l5.0mg/g), begasse fl y ash (17 .0 mg/g), granu lar activated carbon (20.0 mg/g) and sa lvini a molesta Mitchell-based acti vated carbon ( 11 5.06 mglg) II. The data from the Table 2 show that the lower temperature promotes adsorpti on of phenol, wh ich has stronger coloumbic interaction with the adsorbent and adsorbate . As the temperature increased from 10 to 400e, the QO of phenol lost more than 22 per cent of its adsorption. The essenti a l characteri sti cs of the Langmuir isotherm can be expressed in terms of a

3 r----------------------------,

2.5

2

.-. ~ 1.5 '-'

~

~~ I U

0.5

0

0

Sorbent dose : 2 gIL pH : 5.5 Agitation time : 4 h Agitation speed : 200 rpm

50 100 150 200

C. (mg/L)

• : 10°C • : 20°C A. : 30°C

• : 40°C

250 300

Figure 6- Langmu ir adsorption isotherm fo r phenol on acti vated carbon

Table 2 - Langmuir and Freundlich constants and thermodynamic parameters fo r the adsorption of phenol on acti vated carbon

Langmuir Constants Freundlich constants Thennodynamic parameters Temperature

("C) Q" b ? ~g KF lin r2 ~g K"

~G" c (mg/g) (Llmg) (per cent ) (per cent) (kJ /mol)

10 169.80 0.0648 0.9989 5.40 20.8 1 0.4455 0.89 11 18.77 2.843 -2.458

20 156.50 0.03 18 0.9988 3.27 12.23 0.4952 0.9465 12.73 1.98 1 - 1.666

30 146 .1 6 0.01 82 0.9934 5.86 8. 11 0 .5077 0.9736 2.4 1 1.3 17 -0.694

40 132.52 0.0 134 0.9986 2.13 5.38 0.5582 0.9658 9.47 0.808 -0.556

136 J SCIIND RES VOL 61 FEBRUARY 2002

dimensionless constant separation factor or equ ilibrium factor, RL which is defined by RL = I 1(1 +bCo), where b is the Langmuir constant and Co is the initi al concentration of sorbate. RL values obtained (data not shown) at different concentrations and temperatures are between 0 and I indicates favourab le adsorption of phenol on activated carbon.

The linear plots of log qe vs log Cc (Figure 7) show that adsorption of phenol onto carbon also fol lows Freundlich isotherm model. The values of Kr: and I In (Table 2) at 300e were found to be 8. 11 and 0.5177, respectively. Values of 0<lln<1.0 show the favorability of adsorption on activated carbon24

. The ultimate adsorption capacity of activated carbon can be calculated by substituting the required equilibri um concentration in the Freundlich equation . Thus for an equilibrium conc. of I mglL of phenol, I g of the activated carbon can remove 8.1 I mg phenol at 30°e.

Ooe 25 . I For phenol adsorption at 3 on peat ,commercia activated carbon, coal fly ash22 and organocla/6

, the values of Kr: were reported to be 0.46, 0.22, 0.31 and 0.50 mg/g, respectively, which shows that adsorption capac ity of coconut husk based activated carbon is very high .

To judge the fitting of the above two models the adsorption data on activated carbon, a normalized deviation (t.g%) was calculated by the Eq. (6).

N ( ex pt _ calc) 100" \qe.; qe.;

t::.g (percent) = - L expt' ... (6) N ;=1 qe, ;

where the superscripts 'ca lc' and 'expt' are the calculated and experimental values. N is the number of measurements. The va lues of t.g (per cent) are li sted in Table 2. From the results it was found that t.a obtained from Langmuir equation is smaller (t.g =

b .

2.1 - 5.9 per cent) and it is somewhat better than of the Freundlich plot (t.g = 2.4 - 18.8 per cent).

Thermodynamic parameters were calculated from the variations of the thermodynamic distribution coefficient, Ko with change in temperature. Ko for the phenol adsorpt ion was determined by the procedure described by Khan and Singh27 by plotting In qelCe vs qc and extrapolating to zero qe (Figure 8). The standard free energy change t.Go for the interaction of activated carbon with phenol was calcu lated as t.Go = -Rnn Ko ' From the variation of In Ko with

temperature, the standard enthalpy, t.ff' and entropy t.S' changes were computed using Eq. (7).

t::.s tJ f.,JJ"

R RT ... (7)

The plot of In Ko vs liT was found to be linear. The values of t.ff' and t.So were obtained from the slope and intercept of the plot. The negative values of t.S' for phenol removal indicate the feas ibil ity of the process and spontaneous nature of the adsorption. The negative va lue of Mf' (-30.73 kllmol) indicates that phenol-activated carbon interaction is exothermic. The negative value of t.S' (-99.54

• : lOOC

• : 20°C 2.5 .. : 30°C

2 • : 40°C

0,5

Sorbent dose : 2 gIL pH : 5.5 Agitation time : 4 h Agilation speed : 200 rpm

o L-__ -L ____ ~ __ ~~ __ ~ ____ L-__ -J

o 0,5 1.5 2 2,5 3

log C,

Figure 7-Freundlich adsorption isotherm for phenol on act ivated carbon

3

2

.-.. I

~ ~ '-' a .s

-1

-2

a 50

Sorbent dose : 2 gIL pH

•

100

q. (mg/g)

• •

150

• : !O°C

• : 20°C .. : 30°C

• : 40°C

200

Figure 8-Plots of In (qrJCcl vs qc ror the adsorption of phenol on activated carbon

VINOD & ANIRUDHAN et al.: PHENOL ADSORPTION ON ACTIV ATED CARBON-A STUDY 137

llmol K) indicates a greater order of reaction during the adsorption of phenol onto activated carbon and also reflects the affinity of the adsorbent materi al for phenol.

Information concerning the magnitude of the heat of adsorption and its variation with surface coverage can provide useful information concerning the nature of the surface and the adsorbed molecules. The heat of adsorption determined at constant amount of sorbate adsorbed is known as the isos teric heat of adsorption (I1Hx) and is calcu lated using the Clausius-Clapeyron equation .

d( ln CJ Ml x ---

d t RT 2 . . . . (8)

For this purpose the values of Ce at constant amount of phenol adsorbed are obtained from the adsorption isotherm data at different temperatures. I1Hx is calculated from the plot of In Ce vs liT for different amount of phenol adsorption (Figure 9) . I1Hx values are shown in Figure 10 as a function of the amount of phenol adsorbed. As shown in Figure I 0 the isos te ri c heat of adsorption is varied with the surface loading indicating the activated carbon used as an energetically heterogenous surface.

Experiments were carried out to investi gate whether complete phenol desorption was possible from carbon loaded with phenol. Studies in this direction have been carried out in detail with mineral acids, sa lt s, bases, methanol , distilled water (30DC) and hot water (40-70DC) . Hot water at 60DC has been

8 r---------------,

6

Surface loading • :40 mglg .: 60 mglg .: 80 mglg . : 100 mglg

u~ 4

.s 2

Sorbent dose pH Agitation time : 4 h Agitation speed : 200 rpm

OL------JI....-----L--.......L. __ ~ __ ...J

3.1 3.2 3.3 3.4 3.5 3.6

lrr X UP (K-J)

found to be the most effecti ve desorbing agent. Therefore, the process of phenol adsorption on carbon is reversible but slow at 30De. Due to hi gh adsorbability of phenol at lower temperature onto carbon nearly complete desorption of phenol fro m used carbon may require high temperature. On ly 30.7 per cent desorption took place in disti lkJ water, whereas 99.1 per cent desorption was obsen cd at hot water (Table 3) . This is an evidence of phy~ ica l and exothermic nature of adsorption. After two cycles, the adsorption capacity of carbon decreased from 92.7 to 83.3 per cent, whi le recovery of phenol decreased from 99.1 in the first cycle to 92.2 per cent in the third cycle. These results show that the spent carbon can be effectively regenerated for further use by hot water.

Conclusions

The use of coconut husk based steam-pyrolysed activated carbon as low cost adsorbent for the removal of phenol from aqueous solu tions was studied. The carbon seems to be very effic ient and

Or----y----.r----y----.r----r-----,

,-., -o e

-20

~ -40

'-"

=~ <l -60

Sorbent dose pH

: 2 giL : 5.5

Agitation time : 4 h Agitation speed : 200 rpm

-80 L-_____________ --'

o 20 40 60 80 100 120

Surface loading (mg/g)

Figure 10-Variation of tlH, wi th respect to surface loading

Table 3 - Summary of regenerati on st udies

Distilled water Hot water

No. of -------------------------------------­cycles Adsorption D esorption Adsorption Desorpti on

(per cent) (per cent) (per cen t) (pe rcent)

92.7 30.7 92.7 99. 1

2 88 .6 26.8 90.0 96.7 Figure 9- Plots of In Cc of phenol at constant amounts adsorbed as

a function of ( 117) 3 83.3 24.1 85 .6 92.2 -------------------

138 J SCI IND RES VOL 61 FEBRUARY 2002

economical for the removal of phenol from wastewaters. About 92- 99 per cent phenol removal is possible from synthetic aqueous phenol solutions containing 25-50 mg/L. The maximum removal was observed at pH 5.5. The applicability of external mass transfer diffusion and intraparti cular mass transfer diffusion mode ls has been checked and kinetic parameters as a function of initi al concentration of sorbate, temperature, agitation speed, and particle size of the adsorbent were calculated. Adsorption equi librium data fo ll ows, both Langmuir and Freundlich isotherm models. Temperature study reveals that adsorpt ion of phenol is exothermic in nature. The desorption data show that spent adsorbent can be regenerated for further use by hot water treatment. The raw material , coconut husk used for preparing activated carbon is available almost free of cost and involvi ng only e lectri ca l and chemical (HCI) charges. Coconut husk based activated carbon is likely to be considerably cheaper than conventional activated carbon.

References

Perrich J R. Adsorption for wastell'ater treatment (CRC Press. Boca Raton. FL) 198 1.

2 IS 2490 (Part I) Tolerance limit for industria l effluents discharged into inland surface waters: coke oven, Bureau of Indian Standard. New Delhi , 1974.

3 Banerjee S K & Mathew M D. Carbonization of jute stick. an agricultural waste, Agric Waste, 15 (1985) 225.

4 Raji C, Manju G N & Anirudhan T S. Removal of heavy metal ions from water using sawdust based carbon, Indian J Ellg Mater Sci, 4 (1997) 254.

5 Ch ughtai F A, Nisa F. JII ahi A, Hague C & Parveen N. Activated carbon from a tropical carbonaceous materi al J Pure Appl Sci. 6 (1987) 57.

6 Ranouprez A & Avom J, Characterisation of ac tive carbons from palm tree fibres usi ng nitrogen adsorpti on and small angle x- ray scattering, In: Characterisation of porous solids, edited by K K Unger. J Rangaerot and K S W Singh (Elsevier Publi shers, Amsterdam) 1988.

7 Nawar S S & Doma H S, Removal of dyes from effluents using low-cost agricultural by products, Sci To tal Environ, 79 ( 1989) 271.

8 Srinivasan K, Rao PBS & Ramadevi A, Studies on charac teristics of carbon obtained from tamarind nut. Indian J Environ Health, 30 ( 1988) 303.

9 Pollard S J T, Thompson F C & McConnachie G C, MICropOrous carbon from Moringa Oleifera husk for water puri fication is less developed countries. War Res, 29 (1995) 337.

10 Hitchcock S J, McEnaney B & Watling S J, Fibrous active carbon from coir, J Chem Technol Biotechnol, 33 (1983) 157.

II Sankaran N B & Anirudhan T S. Adsorption dynamics of phenol on act ivated carbon produced from Salvinia molesta Mitchell by si ngle-step steam pyrolysis, Indian J Eng Mater Sci, 6 (1999) 229.

12 Warhurst A M, McConnachie G L & Poll ard S J T. The production of activated carbon for water treatment in Malawi from the waste seed husks of Moringa Oleifera, Wat Sci Technol, 34 (1996) 177.

13 Manju G N, Raji C & Anirudhan T S. Evaluation of coconut husk carbon for the removal of arseni c from water. Wat Res, 32 (1998) 3062.

14 Schwarz J A, Driscoll C T & Bhanot A K, The zero point charge of sili ca-alumina ox ide suspensions, J Colloid Interf Sci, 97 ( 1984) 55.

15 Hallum J V & Drushel H V, The organic nature of carbon su rfaces, J Phys Chem, 62 (1958) 110.

16 Nakani shi K & Solamon P H, Infrared absorption spectroscopy (Holden-Day Inc, San Francisco) 1997.

17 McKay G, Otlerburn M S & Sweeney A G, Surface mass transfer processes during colour removal from effl uent using silica, Wat Res, 15 ( 1981) 327 .

18 Urano K & Tachikawa H, Process development for removal and recovery of phosphorous from wastewater by a new adsorben t-2. Adsorption rates and breakthrough curves, Ind Eng Chem Res, 30 ( 1991) 1897.

19 McKay G. Blair H S & Finton A. Sorption of metal ion by chitosan, Immobilization of ions by biosorption , edited by H Eccles and S Hunt (Elli s Harwood, Chi chester. UK) 1986.

20 Streat M. Patrick J W & Camporro Perez M J. Sorpti on of phenol and para-ch lorophenol from water using convelllional and novel activated carbons, Wat Res, 29 (1995) 467.

2 1 Cooney D 0 & Xi Z. Activated carbon catalysed reacti on of phenolics duri ng liquid - phase adsorption. A1ChE J. 40 (1994) 36 1.

22 Swamy M M, Mall I D. Prasad B & Mishra I M, Remova l of phenol by adsorption on coal fl yash and act ivated carbon. Poll Res, 16 ( 1997) 170.

23 Giles C H, McEwan T H, Nakhwa S & Smith D, Studies in adsorption. Part II . A system of classificati on of solution adsorption isotherms, J Chem Soc, 4 ( 1960) 3973.

24 McKay G. Blair HS & Garden J R, Adsorption of dyes on chitin . I. Equilibrium stud ies. J Apply Polym Sci. 27 ( 1982) 3043.

25 Viraraghavan T & Alfaro F D M, Adsorption of phenol from wastewater by peat, fly ash and bentonite, J Hazard Mater, 57 (1998) 59.

26 Dentel S K, Boltero J Y. Khati b K. Demougeot H. Doguet J P & Anselme C. Sorption of tan nic acid, phenol, and 2.4,5-trichlorophenol on organoclays , War Res. 29 (1995) 1273.

27 Khan A A & Singh R P. Adsorpti on thermodynamics of carbofuran on Sn(IV) arsenos ilicate in H+. Na+ and Ca2+ forms. Colloids Surf. 24 ( 1987) 33 .